Tensorflow 2.0：从 tensorflow keras 导入

Question

I can't import anything from keras if I import it from tensorflow. I installed tensorflow 2.0 with pip install tensorflow , and while I'm able to write something like:

import tensorflow as tf
from tensorflow import keras

model = keras.Sequential()

If I try to import Sequential from keras

import tensorflow as tf
from tensorflow import keras
from keras import Sequential

I got Unresolved reference 'keras' . I've looked into every other post I could find and the information is contradictory, some say you have to install keras separately other says you just need to install tensorflow.
So far I've tried:

from tensorflow.python import keras

from tensorflow.contrib import keras

import tensorflow.keras as keras

from tensorflow.keras import Sequential 

Plus a bunch of combination of the above, none of these work.
Sorry if it's a dumb question but I've never struggled so much with a simple import before.

Edit: Additionnal info, I'm on ubuntu 18.04, with Pycharm and a Python 3.6 virtual environment.

Answer:

It is actually a PyCharm bug: Link here: https://youtrack.jetbrains.com/issue/PY-38220 I tried the snippet of code proposed by @AYI here

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten

example_model = Sequential()
example_model.add(Conv2D(64, (3, 3), activation='relu', padding='same', input_shape=(100, 100, 1)))
example_model.add(MaxPooling2D((2, 2)))
example_model.add(Flatten())
example_model.summary()

And actually runs normally despite the warning and error displayed by Pycharm !

Answer 1

Try in this way should help you "from tensorflow.keras.xxx import xxx"

Example of how to import Sequential in tensorflow 2.0:

from tensorflow.keras.models import Sequential

good luck~

---

Here is the Demo:

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten

example_model = Sequential()
example_model.add(Conv2D(64, (3, 3), activation='relu', padding='same', input_shape=(100, 100, 1)))
example_model.add(MaxPooling2D((2, 2)))
example_model.add(Flatten())
example_model.summary()